On-orbit Satellite Power Component Degradation Status Estimation Method

With the rapid development of space technology,more and more satellites are launched into space to perform various tasks.However,the power system provides power for the on-orbit operation of the satellite.Once the power system fails,the satellite will not work normally.As the function of satellite system becomes more complex,the failure rate in orbit increases obviously,so the health of power system is often an important factor restricting the satellite.So the fault diagnosis and health management(prognostics and health management)technology of the satellite power system is effective and reliable health assessment has very important practical significance and application value.Satellite power system were introduced in this paper two parts,the solar cell array,battery development,characteristics and performance,and to study the degradation estimate method based on the health status,the main work is as follows: first of all,according to the environmental factors affecting the sun battery array output power characteristics,puts forward the solar cell array degradation estimation method based on clustering.This method USES clustering algorithm to cluster environmental factors to find the same working conditions of solar array in space environment,so as to evaluate the output power degradation of solar array reliably.At the same time,a mathematical model of solar array degradation based on wave peak current is proposed.In this method,the peak current variation is divided into periodicity and long-term decay to fit the output power fluctuation,and the long-term decay part is used to evaluate the degradation of solar battery array.Meanwhile,the trend of the epitaxial model can also be used to predict the change of the output power.Secondly,for the battery,another important component of the satellite power system,a method based on sample entropy is proposed to predict the health state interval of the battery.Firstly,cut-off voltage and sample entropy were extracted from discharge voltage data as input and capacity data as output.The optimized upper and lower boundary estimation neural network model can effectively reduce the model error and make the prediction interval more accurate.Finally,a large number of experiments and a number of satellites were carried out to compare the proposed degradation analysis method,and it was verified that the proposed method could effectively realize the health analysis of satellite power supply components,which had obvious advantages over the existing methods.